面向直接示教的机器人负载自适应零力控制

发布时间：2018-01-22 21:36

本文关键词： 零力控制自适应控制模型辨识直接示教变负载　出处：《机器人》2017年04期 　论文类型：期刊论文

【摘要】：相对于基于末端多维力传感器的直接示教工业机器人,基于关节扭矩传感器信息的机器人可接触范围大,且安全性更好.但在变负载的复杂工况下,加上受到机器人自身摩擦力、重力、惯性力的影响,示教效率受到制约.为此,本文提出了一种零力控制方法.首先,引入柔性关节机械臂的动力学模型作为被控对象,并分析了该模型中电机摩擦力、惯量以及机械臂重力在直接示教过程中的影响.然后,为精确补偿机器人自身的重力,基于QR分解与最小二乘理论进行参数辨识.而且针对机器人更换末端执行器或抓取物品后模型参数发生变化的情况,提出一种变负载自适应零力控制方法.最后,在自主研发的7自由度协作型机器人平台上进行了实验.机械臂自身重力参数辨识后的模型最大计算误差在关节额定力矩的4%以内,单关节力牵引实验中通过调节零力控制参数,牵引力矩可由约13 N·m下降到约2 N·m.变负载下的机器人牵引实验中,控制器能够在10 s内更新参数.实验结果表明,在负载变化的情况下,该控制方法可帮助操作者轻松地拖拽机器人进行精准、高效的直接示教.
[Abstract]:Compared with the direct teaching industrial robot based on the terminal multi-dimensional force sensor, the robot based on the joint torque sensor information can reach a large range, and the safety is better, but under the complex conditions of variable load. With the influence of friction, gravity and inertia force, the teaching efficiency is restricted. Therefore, a zero force control method is proposed in this paper. First of all, a zero force control method is proposed. The dynamic model of flexible manipulator is introduced as the controlled object, and the effects of motor friction, inertia and manipulator gravity in the direct teaching process are analyzed. In order to accurately compensate the gravity of the robot, the parameters are identified based on QR decomposition and least square theory, and the model parameters change after the robot changes the end actuator or grab the object. A variable load adaptive zero force control method is proposed. Experiments were carried out on the self-developed 7 DOF cooperative robot platform. The maximum calculation error of the model after the identification of the gravity parameters of the manipulator is less than 4% of the rated torque of the joint. In the single joint force traction experiment, the traction torque can be reduced from about 13 N 路m to about 2 N 路m by adjusting the zero force control parameters. The controller can update the parameters within 10 seconds. The experimental results show that the proposed control method can help the operator to tug the robot easily and accurately and efficiently in the case of load change.
【作者单位】：中国科学院沈阳自动化研究所机器人学国家重点实验室;中国科学院大学;
【基金】：基金项目:下一代工业机器人关键技术及系统开发(KGZD-EW-608-1) 新一代机器人控制技术研发(2015106014) 与人共融机器人共性支撑技术(C2016001)
【分类号】：TP242
【正文快照】： 1引言(Introduction)机器人示教是机器人投入使用前的重要步骤,现有的工业机器人示教方法以示教盒编辑示教动作为主.这种示教方法需要操作者对机器人动作逐个调试,组成一整套动作并使机器人重复示教动作.该示教方法使用时繁琐、低效,且对使用者的技能水平要求较高.对于装备大

【相似文献】